1. Convergence in the EU: what role for industrial relations? · the impact of industrial relations...

9

1. Convergence in the EU: what role for industrial relations? Daniel Vaughan-Whitehead and Rosalia Vazquez-Alvarez

1. INTRODUCTION

Convergence in wealth and living standards represents one major objective of the Euro-pean Community construction, as recently evidenced by the objectives of the European Pillar of Social Rights (EC 2017a, 2017b). The aim of this volume is to provide insights into economic and social convergence trends in the EU and to better identify the possible contributions of industrial relations to such trends.

While the literature on convergence to date has focused mainly on ‘economic conver-gence’ – and notably by means of comparing GDP per capita between countries over time – the contributors to this volume look at the convergence of European Union countries (EU28) on a greater number of socio-economic dimensions, as evidenced in each of the independent chapters that comprise the volume.

Besides the contribution of each individual chapter, the aim of this first chapter is to better understand the importance of convergence as one major objective of EU construc-tion and the place it is currently taking in the EC policy agenda (Section 2).

We then present some of the main inputs from the national chapters, which reconstitute their country’s ‘national story’ in terms of convergence, and provide some evidence of the impact of industrial relations on convergence (Section 3). The experts in their differ-ent chapters offer statistical analyses and, when data are sparse, illustrate their national ‘convergence story’ with a number of in-depth case studies, including a number of cross-country studies. This introduction does not aim at summarising convergence among the country studies but rather to provide some analysis of how industrial relations may have helped to explain country trends in terms of convergence in labour and social indicators.

We then try to present some trends in social and economic convergence among the EU28 through a series of social and labour indicators, taking as a basis the European Pil-lar of Social Rights (Section 4). The selection of indicators was decided by the experts’ group who in their national chapters in the volume further illustrate national differences in socio-economic dynamics across Europe. This EU28 comparative analysis is conducted with the aim of identifying trends and convergence among member states, and then of understanding such trends in relation to changes in industrial relations.

This leads us to discuss a number of policy issues in the conclusion (Section 5), which are further explored in the national chapters in light of particular domestic contexts.

2. THE GOAL OF CONVERGENCE AS PART OF EU CONSTRUCTION

Convergence Convergence was one basic objective behind the initial European Commu-nity construction and has motivated closer and closer cooperation and integration.

Daniel Vaughan-Whitehead and Rosalia Vazquez-Alvarez

10

The founders of the EC had planned that common economic interests towards eco-nomic integration would drive more convergence in terms of living standards and then also in terms of welfare states and social policies. Jean Monnet, in his address at the sign-ing of the Treaty establishing the European Coal and Steel Community (ECSC) on 19 March 1951, already emphasised:

The pooling of the production of coal and steel through the creation of a single market of 150 million consumers covering the territory of the six countries, in which all producers will have free and equal access to markets and all consumers a free and equal access to supplies will stimulate technical progress, the increase of productivity and the improvement of the standards of living.1

One of the aims of the Treaty of Rome in 1957 was precisely to reduce ‘the differences existing between the various regions and the backwardness of the less favoured regions’ and to promote ‘harmonious development of economic activities, a continuous and bal-anced expansion, an increase in stability, an accelerated raising of the standard of living and closer relations between the States belonging to it’ (Article 3).

‘Convergence’ was then complemented by the term ‘cohesion’ (economic, social and regional), which acquired a treaty basis in the 1986 Single European Act, and refers to the EU’s ambition to reduce the gaps in development and wealth through a catching-up pro-cess and faster rates of growth for countries and regions at lower levels of development.

It is important, however, to highlight that such a convergence process at EU level was never seen as one that would tend towards parity. Even the Maastricht nominal conver-gence criteria, which put the concept of convergence at the core of the EU’s and national member states’ policy agenda, and applied it to macroeconomic indicators, did not require convergence to a single point, but rather towards a range with upper and lower bounds, which would accommodate differences among member states.

2.1 EC tools aimed at ensuring convergence and coordination

The EC put in place a considerable number of Community programmes, instruments and resources to facilitate this convergence and cohesion process, as reflected in the steady growth of structural and investment fund expenditure as a percentage of the EU budget. The European Charter of Fundamental Social Rights clearly enshrined social objectives as a common goal, while European structural funds were precisely aimed at promoting the convergence of less developed countries and of the most backward regions within those countries. It also led to a widening of the scope of the instruments, not least the Cohesion Fund targeted at member states with relatively low GDP (Treaty on European Union 1992) and to a significant increase in resources available for cohesion policy.

The EC progressively extended the objective of convergence to other areas normally considered as being under national competence. Employment was first recognised as a European policy goal at the Amsterdam Summit of 1997, before being placed in the core of the open method of coordination in the European Employment Strategy of 1998 and further confirmed in the Lisbon strategy in 2000 within the framework of the common

1 Address given by Jean Monnet at the session to sign the ECSC Treaty at the Quai d’Orsay, Paris, 19 March 1951, p. 2. htt-ps://www.cvce.eu/obj/address_given_by_jean_monnet_at_the_session_to_inial_the_ecsc_treaty_paris_19_march_1951-en-5752bf8f-4cfa-4553-bcbf-0cc14fcfa37b.html

Convergence in the EU: what role for industrial relations?

11

leitmotiv ‘for more and better jobs’. Although considered to be ‘soft law’ the open method of coordination – first applied to employment, then to retirement and social inclusion – did encourage national member states towards common goals, exchange of best practices among them and the implementation of National Action Plans monitored by the European Commission. The term ‘coordination’ is very close to the term ‘convergence’ and must be interpreted to mean using common objectives and common tools to facilitate convergence towards a number of common economic and social indicators across EU countries.

2.2 Convergence a priority in EU enlargement waves

The various waves of EU accession – beyond their political goal of integration – were also aimed at bringing countries with a lower level of development (first the Southern countries and then the Central and Eastern European countries) on to the road of conver-gence with EU economic and social standards.

The enlargement waves have increased the challenges facing cohesion policies. With the exception of Austria, Denmark, Sweden and the United Kingdom, all additions to the original six members of the European Union had a GDP per capita below the EU average at the time of joining the group.

This ambition of convergence clearly represented a priority in the EU enlargements of the mid-2000s. Indeed, the enlargements of 2004 and 2007 were unprecedented, not only in terms of the number of new members, but also their relatively low level of GDP per capita. Accession to the EU was also considered by the candidate countries – then the new members of the EU – from Central and Eastern Europe, as well as Cyprus and Malta as a lever for economic catch-up.

EU accession was also accompanied by rapid growth in foreign investment, also seen as an engine for economic convergence. This association between EU membership and convergence was confirmed for these last comers, which have experienced spectacular growth. However, it might be interesting to study whether withdrawal from the EU may lead to different catch-up paths, something that the chapter on the UK investigates by ex-ploring the potential of convergence performance of the United Kingdom after ‘Brexit’.

2.3 A convergence process interrupted by the crisis

The severe economic crisis of 2007–2008 seems to have interrupted what had hitherto been a process of convergence of standards of living among EU member states, some-thing that becomes evident in the statistical analysis in Section 4. The European conver-gence machine stalled – and in some cases even went into reverse – while the financial crisis and the threat to the euro revealed the need for stronger economic governance.

During this crisis period, the common goal of convergence somehow also disappeared from public policies and discourses, with national countries reacting to the crisis in very different and uncoordinated ways. Furthermore, the fiscal consolidation policies applied by the Troika – with the European Commission as part of it – in Greece, Portugal and Ireland led to adverse effects on a number of social indicators: as result, expected social changes that had been previously assumed as common goals among the EU28 suffered a setback, for example, expectations in terms of increased living standards, increased col-lective bargaining and the implementation of statutory minimum wage policies, among others. Besides the difficulties involved in overcoming the crisis, an understanding also


12

emerged that the stability of the single currency required convergence on more than just the monetary and fiscal indicators adopted at Maastricht. ‘It probably requires some prior convergence of labour market policies and characteristics’ (European Commission 2017c). Thus, the aftermath of the financial and economic crisis, and the adverse effects left over in some of the most vulnerable economies among the EU28, has acted as the catalyst to put the convergence concept back at the centre of European policy discourse.

2.4 The European Social Pillar: signal of a renewed process of convergence?

In 2015, the Five Presidents’ report, Completing Europe’s Economic and Monetary Un-ion, stated: ‘the notion of convergence is at the heart of our Economic Union’ (European Commission 2015b). The objective of coordination and convergence was also re-empha-sised in the Europe 2020 guidelines, which included common targets in terms of employ-ment rates, social inclusion and education policies.

The term also gained other meanings and nuances, primarily in the context of govern-ance of the Economic and Monetary Union (EMU). While the term ‘convergence’ was previously used in relation to monetary and fiscal indicators (as a prerequisite of a stable EMU), the trials of the euro in the early period of the financial crisis indicated that con-vergence on only these nominal indicators was not sufficient. The Five Presidents’ report introduced in this context a greater focus on employment and social outcomes. It referred to ‘structural convergence’ and to the need for ‘both more efficient labour and product markets and stronger public institutions’ (European Commission 2015b).

In 2017, the European Commission’s launch of the European Pillar of Social Rights, a new flagship initiative then presented as an EC recommendation, represented an attempt to bring social policies back to the core of the policy agenda and to renew European co-ordination to further promote upward social convergence across the EU, ‘designed as a compass for a renewed process of upward convergence towards better working and living conditions in Europe’. There are 20 principles and rights enshrined in the Pillar (includ-ing ‘Social dialogue and involvement of workers’) structured around three categories: (i) equal opportunities and access to the labour market; (ii) fair working conditions; and (iii) social protection and inclusion (European Commission 2017a).2

‘The logic of the Social Pillar more specifically is a social convergence logic of the economic governance of the EMU’ (Eurofound 2017: 9).

2.5 Industrial relations to serve the convergence process

As emphasised by President Juncker, the EU needs to score ‘a triple A on social issues’.3 In the same direction – and within this revitalisation of social policies – the European Commission set up a new ‘Val Duchesse’ round with European social partners to boost European social dialogue and to call for a more substantial involvement of social partners

2 http://ec.europa.eu/social/BlobServlet?docId=17625&langId=en. A social scoreboard has been established as part of the Pillar to monitor progress on the ground. It will expand and replace the existing scoreboard of key economic and social indicators used in the European Semester of economic policy coordination to better take into account social considerations. This will serve to track trends and performances across countries and to inform policy guidance in the context of the European Semester. See https://composite-indicators.jrc.ec.europa.eu/social-scoreboard/ and also a brochure http://ec.europa.eu/social/BlobServlet?docId=17613&langId=en

3 Statement by President Juncker at the European Parliament plenary session, Strasbourg, 22 October 2014. http://europa.eu/rapid/press-release_SPEECH-14-705_en.htm


13

in EU policymaking and in new legislative initiatives. More emphasis was also given to capacity-building of social partners.

This socio-economic convergence is increasingly seen as necessary for both the sta-bility of the single currency and before any further deepening of the integration process, not least in terms of fiscal governance. Such ex ante social convergence before further development of EMU is in marked contrast to the sequencing of the social dimension in the context of the advent of the ‘single European market’, when better social policy was expected to be the natural consequence of economic integration.

In light of such an ex-ante goal, it is important to understand how far countries have travelled in terms of converging with each other, or diverging, in matters of labour market and social outcomes, and the extent to which the financial crisis impacted on such a goal. This is precisely the objective of each of the chapters that make up this volume, in which national stories help to illustrate the social and economic dynamics in different European contexts. In the next section we try to summarise some of the most important messages of the national chapters in terms of the contributions of industrial relations to convergence. Complementing the national chapters, we offer some statistics on convergence trends and industrial relations in all EU28 member states.

3. TYPES OF INDUSTRIAL RELATIONS EFFECTS ON CONVERGENCE: NATIONAL EXPERIENCES

While studying convergence trends, the different chapters in this volume provide concrete examples of how industrial relations may have contributed to the convergence process so far. Drawing upon the experts’ contributions in the five sub-sections below we highlight five ways through which industrial relations seem to have contributed to convergence national stories.

3.1 Role of social partners in national policy issues

Several examples in this volume highlight the contribution of the negotiation and signa-ture of tripartite national pacts to maintaining peace, while adjusting labour market con-ditions to the macroeconomic context. Industrial relations through such national social dialogue is found to have contributed to convergence.

The chapter on Belgium in this volume shows that the wage moderation on which the social partners have reached agreement has contributed to an increase in employment and also to development of the dual-earner household model, which increased consumption and also made it possible to achieve a number of macroeconomic goals. Various social dialogue mechanisms and institutions, such as extension mechanisms and indexation, have also helped. As shown by the author, this strong concertation model led to real wage growth across the board and – together with extensive redistributive efforts, through both benefits and progressive taxes – helped Belgium in the end to achieve a more equitable spread of the benefits of economic growth than in many other rich EU economies.

Interestingly, as shown in this volume, the Netherlands, despite similarities with Bel-gium, has followed distinctly different trajectories: while the social partners have also embraced wage moderation, with an even greater effect on employment – and thus pro-gress in terms of access to the labour market – this has often taken the form of part-time work and, more recently, temporary work, doubling its incidence, and growth in flexible


14

and insecure contracts, with developing low-paid employment. This increase in tempo-rary work – that according to the author the social partners have failed to contain – leads him to conclude that this became a clear case of divergence, as it has led the Netherlands away from the EU average, which increased only slightly.

By contrast, trade unions in Slovenia played a role in better regulating the use of non-standard forms of employment. The chapter on Slovenia shows how they put the focus on this aspect rather than merely defending the established standard of regular employment. A common case study with Italy in this respect (appended to Chapter 7 on Italy) shows that, despite Italian trade unions’ initiatives to improve the working conditions of non-standard forms of employment and notably temporary work, a lack of consultation with the social partners by the government led to less balanced outcomes.

Some wage moderation was also agreed by the social partners at macro level in Sweden in 2008–2010 to preserve employment stability and limit increases in unemployment. At the same time, Sweden experienced the highest growth rate in Europe in 2014–2017 and has performed better on many indicators of the European Social Pillar. This is explained by the author of the Swedish chapter in terms of a model of flexicurity that has allowed strong external flexibility, while strengthening workers’ employability through a national training system in which the social partners are directly involved, as well as ensuring high wages and working conditions. Social protection – in which the social partners are also involved – has also remained a strong asset with regard to maintaining Sweden’s good record in terms of social and economic indicators.

Similar wage moderation was agreed through a series of social pacts in Ireland that also contributed to economic growth and social peace in the early 2000s. The chapter on Ireland in this volume highlights that convergence was to some extent halted during the crisis years when they had to implement a strong fiscal consolidation policy. However, Ireland also experienced a very rapid economic recovery. The author highlights that so-cial dialogue may have contributed to this. In fact, the series of agreements after 2009, after some interruption of tripartite social dialogue in the crisis years, does nevertheless represent a degree of continuity in social dialogue and illustrate the utility of maintaining social dialogue as a problem-solving response to economic crisis. These agreements have led to some pay cuts and reductions in public sector employment, as well as reforms in work practices and redeployment, but have also helped to provide job security for pub-lic sector workers on permanent contracts. They also provided for pay restoration when macroeconomic and fiscal conditions improved. This would help to explain why Ireland managed to reduce inequalities, thus diverging in a positive manner from the average of the EU, where inequalities increased gradually between 2007 and 2016.

The social partners have also played a prominent role in the regulation of the French labour market, especially in the determination of wages (that have remained dynamic in France even through the crisis) and the application of social rights for workers. They have directly contributed to maintaining high levels of protection through unemployment insurance and to developing new rights (such as accounts and compensation for ‘ardu-ous work’, which play an important role in a period of retirement reforms). The French chapter also explains how industrial relations have helped older workers’ employment and also gender equality, and how these different forms of involvement have contributed to social convergence.

Industrial relations are also found to have contributed to the catching up and social and economic convergence processes of Southern European countries in the 1970s–1980s, when they were initially less advanced, by ‘providing the necessary context for stimulating the labour market, creating a more favourable environment in the workplace, contributing


15

to a fairer distribution of income, as well as by promoting stable prices’, notably through social pacts. One major feature of the Spanish story in this volume is the lack of con-vergence in social expenditure, which, since the mid-1990s, has remained stagnant as a share of GDP during more than a decade of high GDP and employment growth – social convergence being somehow decoupled from economic growth – and led to an anaemic welfare state. As a result, increased inequality and poverty in Spain could be explained by this lack of growth in social expenditure. In such a context of social spending stagnation, the trade unions in Spain, within a larger campaign on social protection, contributed to some important improvements made in recent decades, such as the creation of a system of non-contributory universal pensions for retirees with no access to standard contributory retirement pensions in 1990, or, more recently, to the successful promotion, during the crisis, of an extension of unemployment assistance for the long-term unemployed without unemployment benefit protection. They have recently been deeply engaged in the debate on the need for a guaranteed minimum income to fight poverty.

Similarly in Portugal, important agreements between the social partners have allowed several reforms of labour law, but also social recognition of the need to tackle certain structural deficits in the Portuguese labour market. The chapter on Portugal shows that they notably agreed on key directions for convergence, such as the need to go beyond competition based on low wages, the recognition of skill shortages and the need to invest in both education and training, and also to maintain a healthy social security system. They were present in the recent public debate on the issues of ageing and fertility, and worked together to promote gender equality and reconciliation of work and family life.

The involvement of the social partners was found to be more limited in the Baltic states and thus to exert less of an influence on social and economic outcomes. Industrial relations around minimum wage fixing, however, seems to have played a role in limiting the growth of inequalities in Estonia. The involvement of the social partners as regards the minimum wage was also found to be determinant in a number of other EU countries (Ireland, Slovenia, Germany).

The role of the social partners is not limited to the regulation of the labour market strictly speaking, but can extend to influence other national policy issues.

For instance, trade unions in Slovenia, which are involved in tripartite debates, man-aged to counter the plan for a flat tax rate in the mid-2000s, which might have had adverse implications on equality and convergence towards better inequality outcomes.

Some chapters in this volume show that industrial relations (notably through social partners’ involvement in national policies) can help to improve education. In Sweden the social partners have helped to promote occupational mobility policies and the develop-ment of skills over the life course, with constant skills upgrading and a low percentage of low paid workers. As a result, Sweden today has the highest levels of job quality in the EU (see details in chapter on Sweden). In France, social partners are also directly involved in the definition and management of lifelong learning, while job quality is offi-cially among the goals of some trade unions (see chapter on France in this volume). There is also a statutory obligation to have some collective bargaining on the quality of working life (every four years). At EU level, the European Trade Union Confederation (ETUC) has also formulated clear goals with regard to job quality.

The story of the United Kingdom is also interesting, showing in terms of convergence that while that country has been excellent on certain indicators, such as access to the la-bour market, and thus has been performing as a good job-generating machine, but with more modest results in terms of delivering decent work for all, especially for women. This situation is explained in part by its weak mechanisms and structures for social dialogue,


16

and partly because trade unions do not play a substantive role in designing and shaping economic, social or labour market policy. This institutional context is expected, accord-ing to the author, to be further diluted after Brexit. The case study in the UK chapter also shows that the social partners have been involved, for instance, in campaigns against bogus self-employment at national and sectoral level – notably in in the logistics and transport sectors – but that the gains of such campaigns have not been widely diffused because of an erosion of social dialogue institutions.

3.2 Value-added provided by collective agreements

In Sweden, the bipartite nature of some labour market regulations, coupled with high union density and a high coverage rate of collective bargaining, have created a favourable institutional environment for the emergence of negotiated compromises aimed at balanc-ing flexibility and security in the labour market, as stressed by the chapter on Sweden. This has led to innovative arrangements on wages, flexible working hours and also vo-cational training at both industry and enterprise level. At the same time, Sweden is one of the few EU countries which has experienced an increase (by 3 per cent) in the labour share in GDP between 1993 and 2015, in contrast with most other EU countries, in which the labour share has declined continuously.

At the same time, there are examples of collective agreements at industry or enterprise level that have extended and improved the statutory universal basic protection in the form of higher replacement/compensation rates over a longer period of entitlement, for example regarding unemployment and sickness benefits, parental and training leaves and pensions (occupational pensions).

Lifelong learning schemes or arrangements and also working time accounts decided by collective bargaining have also contributed to enabling the upgrading of skills and in-creasing opportunities for women and young people, as in Sweden, something that is also being attempted now in France. In Germany innovative agreements have been concluded by the social partners allowing a more variable distribution of working time over the year and also through the various phases of working life, which has contributed to the flexibil-ity required for employment, growth and convergence. Many collective agreements have also reduced the level of overtime in order to make jobs available for younger people, particularly trainees, thus also responding to youth unemployment.

The chapter on Portugal highlights how the social partners have started to negotiate a number of collective agreements that include arrangements to reconcile work and family life; by this means, among other things, they hope to contribute to the national goal of increasing fertility rates.

Similarly in Spain, industrial relations, with an automatic extension of collective agreement and a wide collective agreement coverage, even in a context of low trade un-ion density, have certainly contributed to limit wage inequality before the crisis hit (see chapter on Spain).

In Ireland, the resumption of collective bargaining across most of the public sector, and the development of pattern bargaining in the private sector in the context of economic recovery, suggests some continuity, and a degree of coordination in social dialogue.


17

3.3 Convergence as a benchmark for collective bargaining

Collective agreements in some EU countries define wage increases based on monitoring of wage-setting practices in neighbouring countries. An interesting example is the Bel-gian framework, introduced in 1996 to preserve the country’s competitiveness in EMU by keeping wage development in line with its main trading partners. A national body regularly reports on the evolution of Belgian competitiveness relative to its main trading partners, such as the Netherlands and Luxembourg. These reports are used by the social partners to fix a wage norm for the next round of wage negotiations. Although the norm is only a non-binding guideline, it has generally been respected. The system seems to have worked fairly well: it has kept untouched the wage formation and bargaining system that existed prior to euro accession, but had made the behaviour of the social partners compat-ible with membership of the euro area (Sapir and Wolff 2015).

While the Belgian system cannot be exactly duplicated by other euro-area countries be-cause they have different wage formation and bargaining systems, all euro-area countries could establish a mechanism to ensure that, although operating within their own system, the behaviour of the social partners and the outcome of their wage negotiations would be compatible with euro-area membership in terms of competitiveness and employment.

In June 2015, the Five Presidents’ Report followed up on this idea and proposed to set up a euro-area system of ‘competitiveness authorities’. As a follow-up to the report, in October 2015 the Council adopted a Recommendation requiring euro-area member states to establish national competitiveness boards. This process might be more effective, how-ever, if tackled with the social partners. It could also be extended to countries outside the euro area to facilitate convergence within the larger EU.

Similar to Belgium, in Ireland in 2011 the Services Industrial Professional and Techni-cal Union (SIPTU) agreed a modest pay rise of 2 per cent directly with the relevant com-panies, in a so-called ‘2 per cent strategy’ that took into account wage rises in Germany as well as the European Central Bank target of 2 per cent inflation. By 2014, the ‘2 per cent strategy’ had delivered over 220 collective agreements, covering more than 50,000 workers (see chapter on Ireland). These two examples directly show how industrial rela-tions can help in regulating wages and thus also employment to achieve convergence. This requires that institutions that monitor competitiveness be embedded in social dia-logue and that distributive concerns be mainstreamed in the monitoring of competitive-ness (Vandenbrouke 2017).

To better contribute to convergence, industrial relations should also be able to operate across borders through, for instance, International Framework Agreements (IFA). While such IFA have focused mainly on relocations and redundancies, more studies should be carried out to measure their potential impact on convergence. While the monetary union calls for integrated competitive markets for goods and services and cross-border mobility of labour, this in turn entails a social corollary. In this respect, international framework agreements (Papadakis 2011) and also European works councils (De Spiegelaere and Jagodzinski 2015) have been shown to contribute to convergence, or at least to avoid divergence in working conditions.

3.4 Industrial relations as a factor of resilience

The chapters in this volume also provide examples of industrial relations’ role in induc-ing a more balanced sharing of the costs of the crisis in terms of both working conditions


18

and inequalities at work. In Sweden, innovative practices and solutions were developed by the social partners in the 1990s to deal with job losses and structural changes. Agree-ments covering three-quarters of the labour force have been negotiated to help victims of restructuring to find new jobs rapidly. In Ireland, the Haddington Road Agreement (2013–2018) included pay cuts, but also provided for the restoration of these pay cuts in 2016. The agreement also led to some flexibility to allow for redeployment, work-sharing arrangements and workforce restructuring (see chapter on Ireland in this volume). In Germany, the negotiation of collective agreements on reduced working hours has also made it possible to avoid employment cuts. In part as a result of these different adjust-ment mechanisms negotiated through social dialogue, these countries (with the possible exception of Ireland, where unemployment increased from 4.7 per cent in 2006 to 14.7 per cent in 2012) have experienced only very modest increases in unemployment during the recession. By contrast, in countries where the low presence of trade unions and a lack of social dialogue impeded the pursuit of such mechanisms – as in the three Baltic states – the bulk of adjustments have involved employment cuts, with a very rapid increase in unemployment rates.

3.5 Extending the role of the social partners to new issues

The various chapters of this volume show how industrial relations help – notably through collective agreements – to better regulate the increase in non-standard forms of employ-ment. The case study comparing the Belgian and Dutch experiences shows how a stronger role for the social partners and more formal institutions of social dialogue have allowed Belgium to limit the growth of temporary contracts – which affect 9 per cent of Belgian employees compared with 20 per cent of Dutch employees – and also the association of temporary contracts with part-time work. Similarly, the resilience of trade union mem-bership in Belgium (it has been eroded somewhat in the Netherlands) and the fact that it is distributed more equally among the population – for instance, Belgian women have a union density three times higher than in the Netherlands – have probably contributed to a very low gender pay gap (3.3 per cent), only a quarter of the gender pay gap in the Neth-erlands. Women are also more likely to be the main household earner in Belgium than in the Netherlands. Women’s access to more employment and also to better wages are im-portant for boosting the dual-earner model and for the growth and stability of the middle class, with all the implications of this for social and economic convergence. Certainly, the social partners should also investigate how to improve coverage and protection of young workers. The chapter on Ireland suggests that it may have started to diverge from Europe because its young people have been particularly hit and young workers are subjected to low employment standards.

The case studies in this volume analyse how the scope of industrial relations could be extended to other important areas, such as migration and also foreign direct investment, which obviously may influence both social and economic convergence. For example, in Sweden collective agreements on work practices/work experience schemes and Vocation-al Introduction Employment have targeted young newly arrived migrants lacking relevant experience. Similarly, the case study in the Spanish chapter shows how trade unions in Spain have, from the beginning, embraced the integration of immigrants and contributed to promoting their equal treatment in terms of integration and working conditions.


19

Table 1 Industrial relations – Possible effects on upward convergence

Industrial relations elements / mechanisms

Effects on social convergence Effects on economic convergence

Involvement of social partners to define public sector policymak-ing, education, social services

Can help young people (through education) and women (through better public services) to inte-grate better in the labour marketReduces inequalities (Belgium, Ireland, Sweden)

High participation of young people and women contributes to economic growth

Involvement of social partners in social security fund management or in pensions schemes

Ensures equality in the process (Belgium, France)

Can help in ensuring sustain-ability of the schemes while also reducing deficits

Involvement of social partners in labour market reforms

Can generate more employment growth, and also ensure better protection of non-standard forms of employment (Slovenia)

Can ensure both flexibility and security with employment and productivity gains

Involvement of social partners in training policies

Promotes skills and professional opportunities (France, Sweden)

Contributes to skills upgrad-ing and specialisation in higher value added products and trade (Germany, Sweden)

Tripartite negotiations (and pacts) on wages or incomes policy

Can ensure social peace in pe-riods of wage restraint (Nether-lands)

Can promote economically sustainable wages and incomes policy, and avoid economic dis-ruptions due to industrial unrestMay contribute to wage modera-tion and enhanced competitive-ness (Italy, Ireland and others)

Bipartite or tripartite negotia-tions on the minimum wage

Can reduce wage inequalities between top and bottom and between men and women at the lower end of the wage distribu-tionCan ensure that minimum wages are set at a balanced level, tak-ing into account the needs of workers and their families and economic factors. Increases ownership and effectiveness of minimum wages. Can also reduce migration (con-tributed to net positive migration in Estonia)

Can help to maximise the economic benefits of mini-mum wages (such as aggregate demand or economic benefits of reduced inequality), and mini-mise potential costs Can stimulate investment in in-novation and higher value added segments

Collective agreements Collective bargaining can reduce inequalitiesCan help promote fairer working conditions (example of working time in Sweden)Can help improve working conditions of non-standard forms of employment (higher work-ing hours for part-time workers; better protection for temporary workers or self-employed etc.)

Can help stimulate consumption and thus also middle class stabil-ity and growth (Belgium, France, Sweden)


20

4. CONVERGENCE, THE EUROPEAN PILLAR OF SOCIAL RIGHTS AND INDUSTRIAL RELATIONS

So far this chapter has explained how the various national stories included in this volume assess the possible role of industrial relations in explaining some of the dynamics of so-cial indicators. Whereas the individual chapters provide national analyses, the European Pillar of Social Rights (EPSR), as one of five key policy areas of the European Union, calls for building a more inclusive and fair European Union where all member states come closer (or converge) on a set of 20 principles and rights.4 Based on this call, a natural question to ask is how far apart, or close together, the 28 member states currently

4 The other areas are Economic Union, Financial Union, Fiscal Union and Political Union. The European Pillar of Social Rights refers to 20 principles and rights that include the right to fair wages and the right to health care, lifelong learning, better work–life balance and gender equality, among others. With a call to member states to commit to a common goal in matters related to such principles, the EU stands for cohesion of these rights among all its citizens in a fast-changing world (www.europa.eu/commis-sion).

Table 1 Industrial relations – possible effects on upward convergence (cont.)

Industrial relations elements / mechanisms

Effects on social convergence Effects on economic convergence

Collective agreements (cont.) Can help reduce unsocial hours (working evenings, nights or Sundays) thus helping workers to reconcile work and family lifeCan help in adaptation to new forms of jobs/work and improve their working conditions and protection (example of Uber jobs in Slovenia)Can promote better working con-ditions for migrants (Spain) Can help education and training (training accounts, lifelong learn-ing arrangements)Can help reconciliation of work and family life

Can stimulate growth of newly emerging sectors and new types of job

Enhances skills upgrading and the shift to higher value added jobs (Sweden)

Collective agreements in periods of economic slowdown

Can help reduce layoffs through reduced/shared working hours (Germany, Austria, France)Wage moderation agreements to limit wage growth and employ-ment adjustments (Sweden)Can help in corporate restruc-turing, finding the best balance between flexibility and security (Sweden)

Helps to minimise unemploy-ment rises (Germany, Austria, France etc.)Helps to keep qualified em-ployees within companies and facilitates recovery after crisisContributes to maintain ag-gregate demand and growth (or limit recession)

Collective agreements based on monitoring of neighbouring countries’ practices

Helps alignment on wages and working conditions (Belgium)

Helps to maintain competitive-ness and employment

European works councilsInternational framework agree-ments

Can help to avoid social dump-ing by better aligning working conditions between countries

Can prevent strikes and thus avoid interruptions of activity and falls in productivity


21

are with regard to achieving these 20 principles and rights? In this volume the focus is on the sub-set of these principles and rights that have a direct bearing on the labour market and related socio-economic and institutional outcomes. Moreover, and as highlighted by the national chapters, industrial relations play a key role in the pursuit of common goals and therefore the question naturally arises, what has been the role of industrial relations in the current status of EU28 member states in matters of the labour market and related socio-economic outcomes?

4.1 How has convergence been defined?

Convergence is by definition a long-term phenomenon, despite the fact that economic picks and throws can identify convergence and divergence from some desired goal within short periods of time. In the analysis that follows, convergence is measured over a 17-year period, from 2000 to 2016. In the case of the European Union, and given its rela-tively short history as a unified group, this 17-year period can be considered indicative of convergence in socio-economic outcomes. This is because the period takes into con-sideration the waves of accession years and also covers the pre- and post-financial crisis periods, thus allowing us to capture how it impacted on the aggregate dynamics of today’s EU28 member states. Having said that, most of the other chapters in this volume rely on additional national statistics and cover a period from 1990 or even earlier.

Discussions within the group of experts led to a definition of convergence on the basis of two main criteria. First, because the aim of the EU is to improve the economies and well-being of the population our group of experts defined convergence as a trend towards an improvement of socio-economic indicators. This means that over time the average of the EU should improve in tandem with the progress of individual countries. This does not mean, however, that countries should have as a target to maximise or minimise all indica-tors, given the different nature of the indicators (Table 2 develops this point further).

Within this upward movement, convergence also implies a catching-up process of less advanced countries on those indicators, allowing them to get closer to the average and also somehow closer to the best performers.

The second main feature in our definition of convergence is a lower dispersion among EU countries, while divergence would represent an increased dispersion.

Convergence and divergence movements can be classified in one of four possible sce-narios, as described in Table 2. In this table, both convergence and divergence can be up-wards or downwards. Upwards (downwards) convergence occurs when the values of the indicator are coming together and, on average, each country is doing better than in previ-ous periods; under such circumstances we would observe the EU28 average increasing (decreasing). Upwards (downwards) divergence occurs when the values of the indicator are moving apart, with some countries scoring higher (lower) values compared with the rest – and with previous periods – and above (below) the EU28 average.

The optimal scenario represents all EU countries experiencing a lower dispersion and also improving on their socio-economic indicators with all countries attaining a better outcome in their labour market and social indicators. In this case the overall EU average would increase (Scenario 1).

The achievement of social cohesion as desired by the European Pillars of Social Rights would be associated with this upwards convergence scenario. Such movement would reflect that all EU28 countries are, in tandem, following a positive path, where countries with weaker outcomes in terms of labour and social indicators are improving compared


22

with previous periods and at a faster rate, while countries already exhibiting good labour market and social outcomes continue to do well or even better than in previous periods. In Table 2, upward convergence is described in the top-right quadrant (Scenario 1). The other three scenarios in the table show possibilities that gradually move away from the preferred upwards convergence outcome.

There could be a second possible scenario, with all countries also improving on socio-economic indicators (with then also an increase in the group average), but leading not to a lower but to an increased dispersion because certain countries continue to progress more quickly than others (Scenario 2).

Although less desirable than Scenario 1, this last outcome would still be more ben-eficial for all countries than another scenario that could involve increased dispersion, alongside a general deterioration of socio-economic indicators, thus leading to a decrease in the EU average (Scenario 3). In this case, some or all countries are experiencing a weakening in the underlying labour and social indicators, while the difference between the countries with lower outcomes (for example, a higher unemployment rate and lower social protection spending) moves further apart from countries experiencing relatively better outcomes.

Another possible case would be convergence between EU countries, not towards im-provement, but rather towards a deterioration of socio-economic indicators, which would imply a decrease in the EU average (Scenario 4).

4.2 Overview of convergence using a selection of indicators

One way to answer the above questions is to do a comparative analysis of the 28 coun-tries using a selection of indicators that capture the principles and rights of the European Pillar in relation to labour market outcomes and industrial relations. An inspection of the 20 principles endorsed by the EPSR helps in the selection of indicators to measure the three dimensions targeted by the 20 principles: opportunities and access to labour mar-kets, fair working conditions and inequality and social protection. Table 3 shows these dimensions and corresponding indicators, all of which are consistent with the indicators used in the national chapters presented in this volume. In fact, a review of the national

Table 2 Convergence scenarios in the EU

Progress on socio-economic indicators (EU average in-creases)

Deterioration in socio-eco-nomic indicators (EU average decreases)

Convergence (lower dispersion)

Scenario 1: Upward convergence Progress of all countries with less advanced countries also catching up

Scenario 4: Downward conver-genceAll countries may evolve to-gether towards a deterioration of their indicators

Divergence (higher dispersion) Scenario 2: Upward divergence (dispersion alongside increased performance) Progress of all countries on in-dicators but some countries con-tinue to progress more quickly than others

Scenario 3: Downward diver-gence (Dispersion alongside decreased performance) Dispersion because some may evolve towards a deterioration while the others progress or remain stable


23

chapters would suggest adding a fourth dimension to capture economic performance and redistribution. Table 3 shows 20 indicators that could in principle be compared to shed light on whether the EU28 is converging or diverging with respect to social and economic dimensions proposed by the EPSR. Annex A provides more details on each of the indica-tors listed in Table A1.

The proposal of a comparative analysis of the 28 countries raises two problems: first, jointly analysing 20 indicators is a complex task and potentially misleading if the pur-pose is to understand ‘overall’ convergence or divergence among the EU28. The second problem is that of making a value judgement on the direction that each indicator in Table 3 would have to follow in pursuit of convergence among member states while ‘deliver-ing new and more effective rights for citizens’ – which is what the European Social Pillar prescribes. To highlight this point, Table 3 classifies the indicators in two columns: less controversial and more controversial. For example, it is not controversial to think that member states all aim at the lowest possible unemployment rate, therefore, we could con-clude there is positive convergence in matters of unemployment between the EU28 if we

Table 3 Indicators covering the areas of principles and rights underlined by the European Pillar of Social Rights, by dimensions, as structured by the European Commission

Dimensions Areas Least controversial indicators Most controversial indicators1. Equal oppor-tunities and Access to labour market

Access to labour market

Employment rate, unemploy-ment rate, youth unemployment rate, long-term unemployment rate

Opportunities in the labour market

Participation in training (at least in the past four weeks) as a percentage of dependent workers; participation … if level of education below sec-ondary; … if level of education secondary; ... if level of educa-tion higher

2. Fair working conditions

Wages Wage in 2011 PPP constant val-ues; growth in average wages of full-time employees

Kaitz Index (minimum wages as proportion of average wage)

Employment contract

Involuntary part-time among all part-time employment

Proportion of workforce with temporary contracts

3. Social protection and inclusion

Social protection

Per capita spending on social protection

4. Economic performance and distribu-tional outcomes

Inequality Share of household income in the D1 decile; and of house-hold income in the D10 decile; ratio between total share of household income in D2 to D4 deciles to total share of household income in the D5 to D9 deciles

Economic performance

GPD per capita growth (from GDP in constant 2011 PPP values)

Labour income shares


24

observe that over some period of time these countries are, on average, moving together (converging) towards lower unemployment rates. This would be an example of upwards convergence that hints at social improvement for Europe’s citizens. On the other hand, it would not only be controversial but also undesirable to prescribe that countries fol-low policies of ‘100% permanent contracts’, even if such hypothetical movement, from a quantitative point of view, was also associated with upward movement (convergence) among the EU28.

Despite the two abovementioned problems, coming up with a monitoring and evalua-tion tool that checks for EU28 convergence under the European Pillars remains a poten-tially desirable goal: the tool would highlight the objectives pursued by the Pillar while helping policymakers to make timely and objective assessments of the direction taken by the pursuance of common goals. Meanwhile, and considering that the work in this vol-ume attempts to convey a message for the EU28, it is reasonable to explore quantitative tools that could summarize information available for all 28 economies, over a period of time sufficient for capturing convergence, and in light of the goals implied by the EPSR. This consideration is what drives us to finish this chapter with estimates from a summary measure constructed with the indicators representing the EU28 European Pillar of Social Rights; the summary measure is presented not as an ultimate instrument, but to show the potential of such a monitoring tool. Ultimately, the summary measure is also informative in this context because it provides information from countries in the EU28 that are not present in the national chapters.

The measure proposed implies a unification of the 20 indicators proposed in Table 3 under a single metric, therefore reducing a 20-dimension analysis into a single one: this already takes care of the first problem mentioned above. We call this summary measure a Labour and social summary indicator (LSSI). The technique employed for the purpose is a slight modification of other summary measures, such as the Human Development Index (HDI, UNDP) or the Better Life Index (BLI, OECD). In essence, the proposed measure leads to a summary value that helps us to easily inspect convergence (or divergence) among the EU28 member states with regard to the underlying indicators described in Table 3. Annex 2 provides the basic details leading to the LSSI.

Using the LSSI, the EU28 can be classified into four groups based on a combination of the following two outcomes: whether their LSSI value is higher or lower than the EU28 median in 2016, on one hand, and whether their LSSI value increased or decreased between 2000 and 2016, on the other. This provides the mapping of the EU28 countries presented in Table 4.

We further investigated this issue through a more in-depth analysis of convergence that consisted of plotting the distribution of the estimated LSSI over time.

Figure 1 shows changes in the distribution of the Labour market and social indica-tor (LSSI) measured for each of the EU28 countries, for the period 2000 to 2016: a few years have been selected for clarity of illustration.5 The distribution of the LSSI for each selected year has been summarised into a box-plot, a simple device that highlights the median of the distribution (the thick horizontal line inside the rectangles), the 25th and the

5 The distribution of the LSSI for each selected year has been summarised into a box-plot, a simple device that highlights the me-dian of the distribution (the thick horizontal line inside the rectangles), the 25th and the 75th percentiles of the LSSI distribution (the edges of the rectangles), the extremes (upper and lower lines) and outliers (dots). A vertical stretching or shrinking of the rectangle shows (respectively) divergence or convergence among the LSSI values (that is, among the EU28) over subsequent periods of time, while an increase or a decrease of the median values over time (within those rectangles) shows (respectively) an upward or a downward movement of the values of the LSSI.


25

75th percentiles of the LSSI distribution (the edges of the rectangles), the extremes (up-per and lower lines) and outliers (dots). A vertical stretching or shrinking of the rectangle shows (respectively) divergence or convergence among the LSSI values (that is, among the EU28) over subsequent periods of time, while an increase or a decrease of the median values over time (within those rectangles) shows (respectively) an upward or a downward movement of the values of the LSSI.

According to these measurement criteria, the dynamics of the box-plots over the pe-riod 2000 to 2016 show three clear periods: in the first period, prior to the global financial crisis, the EU28 was moving towards convergence in matters underlying the European Pillar of Social Rights, as captured by the LSSI: a higher median in the boxes showing higher values of the LSSI positively affecting more individuals in the EU28 countries and more concentration (smaller boxes showing lower dispersion) of the values: this could be described as a situation of upward convergence (or Scenario 1) among the EU28.

In the second period, during the financial crisis (2007–2010), the positive convergence achieved just before the crisis was lost by the time countries were coming out of the crisis after 2009: the LSSI shows a downward movement so that on aggregate it seems that the EU28 had declined in terms of the value of the indicators that underline the European Pillar. The value of the LSSI had also widened significantly compared with the year 2000 or compared with immediately prior to the crisis. In 2009 countries were in fact moving apart from each other, with the distribution of the LSSI in that year widening in both di-rections compared with previous periods (downward divergence or Scenario 3).

The third period starts in the aftermath of the crisis; comparing the year 2009 with 2012 shows that divergence among the EU28 countries has continued in the two direc-tions: countries resilient to the crisis in terms of the indicators underlying the European

Table 4 Mapping of EU28 countries according to values of LSSI in 2016 and changes in LSSI between 2000 and 2016 (median wage based on unweighted value)

INCREASE in value of LSSI between 2000 and 2016

DECREASE in value of LSSI between 2000 and 2016

Countries with LSSI score in 2016 ABOVE or at the medium value of the EU28 score in 2016

AustriaBelgium

Czech RepublicDenmarkFinlandFrance

GermanyLuxembourg

MaltaSloveniaSweden

IrelandThe NetherlandsUnited Kingdom

EU28

Countries with LSSI score in 2016 BELOW the medium value for the EU28 score in 2016

BulgariaEstoniaHungaryLatvia

LithuaniaSlovakia

CroatiaCyprusGreeceItaly

PolandPortugalRomania

Spain


26

Pillar continued to improve in the years leading to 2012, while in the same period coun-tries below the median continued to decline. The LSSI in 2016 shows a slightly higher value compared with 2012, but over the four years the movement was marginal, and catching up to the levels of ‘convergence’ observed in 2006 was yet to be achieved. The distribution in 2016 shows a clear divergence between countries above and below the median, especially when compared with the distribution of the LSSI in the year 2000 (Scenario 3 type of ‘downward divergence’).

Putting all this in numerical terms, and excluding outliers, the (weighted) variance among the values of the index in 2005 was estimated at 0.012; in 2012 this had increased to 0.021 – that is, by 75 per cent. In 2016 the variance had declined to 0.018 but this is still 50 per cent more than the variance in 2005, when countries showed more convergence in their performance with regard to labour market and social dynamics.6

In brief, these results show that there was indeed convergence before the crisis that was then interrupted by the crisis, after which more dispersion prevailed. Overall we observe a long-term divergence rather than convergence between 2000 and 2016.

4.3 What relationship between industrial relations and convergence?

Union density and collective bargaining coverage are two indicators often considered to reflect the outcome of industrial relations institutions – for example, the outcome of multi-employer bargaining processes, extension mechanisms or tripartite pacts and agree-ments, to mention only a few. In this volume the national chapters highlight the impor-

6 An estimation of the same dynamics of the LSSI but using unweighted values shows similar results, with the median following the same pattern over time, although at slightly different levels. In general, the unweighted median is slightly below the weighted one, particularly in the post-crisis period. This shows that among countries with larger populations – in particular France and Germany – the LSSI scores higher. The convergence and divergence patterns, however, remain similar between weighted and unweighted estimates.

Figure 1 Changes in the distribution of the Labour and Social Summary Indicator (LSSI) over time, EU28, 2000–2016 (selected years)

Note: ILO estimates. The years in dark grey are the crisis years. The figures show the weighted distribution of the LSSI for the EU28.

0

.1

.2

.3

.4

.5

.6

.7

.8

.9

1

Scor

e va

lues

of t

he L

SSI

2000

2005

2006

2007

2008

2009

2012

2016


27

tance of industrial relations for strengthening labour market and social dynamics, with evidence and data from each of the individual countries and from the various case studies.

The question is, to what extent can the evidence from the LSSI, in view of the evolu-tion of industrial relations, provide further backing for the national stories, and evidence for the aggregate EU28? We start by examining the relationship between collective bar-gaining and union density to understand the complementarity between these two vari-ables in capturing the evolution of industrial relations among the EU28 member states. Figure 2 compares the relation between collective bargaining coverage and union density in 2016, whereas Figure 3 shows the relation between changes in collective bargaining coverage and changes in union density during the past 17 years (2000 to 2016).

In particular, Figure 2 shows a strong relation between the two outcomes, suggesting that both union density and collective bargaining coverage probably result from the same enablers of industrial relations at country level. Thus, countries that have strong coverage are also those with strong union density (for example, Scandinavian countries, Belgium, Malta, Cyprus, Italy and Luxembourg), whereas countries with low coverage are associ-ated with low unionisation (for example, Estonia, Latvia, Lithuania or Bulgaria). There are some outliers that show strong collective bargaining coverage (for example, France and Austria), but relatively low levels of unionisation, but the evidence for the opposite case are rather weak, that is, we do not see countries with high union density similar to Scandinavian ones that, at the same time, have weak collective bargaining coverage. Thus, despite a few outliers – with France serving as an example, mostly due to its indus-trial relations institutions and practices – a correlation coefficient of 49.3 would suggest that collective bargaining is seemingly related to union density, and where the latter is high the tendency is to have higher levels of collective bargaining coverage.

At the bottom left end of the diagram is a group of countries that has very low collec-tive bargaining coverage and trade unionisation, including the three Baltic states (Estonia,

Figure 2 Collective bargaining coverage and trade unionisation, by country and EU28, 2016

AT BE

BG

HR

CYCZ DE

DK

EE

FI

FR

GR

HU IE

IT

LV

LT

LUMT

NL

PL

PT

ROSK

SIES

SE

GB

EU28

0

20

40

60

80

100

120

0 10 20 30 40 50 60 70

Col

lect

ive

Barg

aini

ng C

over

age,

201

6

Union Density, 2016


28

Latvia and Lithuania with less than 10 per cent on both rates), Poland, Slovakia, Bulgaria and Hungary.

The question is, how did the two variables, collective bargaining and union density, move over the period 2000 and 2016? And can such movements shed light on the relation between the two variables that we observe in Figure 2? Figure 3 shows that falling union density is associated with falling collective bargaining coverage, so it is not surprising to see that in 2016 the two variables had a positive correlation. Figure 3 also shows that only a few countries (Finland and Malta) experienced an increase in collective bargaining coverage between 2000 and 2016 – because almost all countries were at best at, but gen-erally lower than the zero level – and all countries (except for Belgium, France and Italy) experienced a 10 per cent or more decline in union membership, with extreme cases such as the Czech Republic, Estonia, Latvia, Romania and Slovakia, where the decline was above 55 per cent. The figure indeed shows an overall weakening of industrial relations among the EU28, with very few countries (eight out of the 28) showing almost no change over the 17 years in collective bargaining coverage, while all other countries declined. Figure 3 in fact seems to show a divide between the above countries and another group of countries – generally older EU member states – that experienced some erosion, but not to the same extent.

In fact Figure 4 completes the trio of illustrations to show that countries with higher levels of collective bargaining coverage in 2016 are the ones that experienced a lower decline in union density over the 17 years (2000 to 2016).

Figures 2 to 4 show the weakening of industrial relations among the EU28 leading up to the year 2016 using two well-known outcomes of labour market institutions in the EU. It is interesting to further explore the evolution of industrial relations over each of the 16 years for which we have data: because both collective bargaining coverage and union density show similar trends, we select the former to clarify the analysis.

Figure 3 Collective bargaining coverage and trade unionisation, by country and EU28, 2000–2016

AT BE

BG

HR

CY

CZ

DE

DK

EE

FI

FR

GR

HUIE

ITLV

LT

LUMT

NL

PL

PT

ROSK

SI

ES

SE

GB

EU28

‐90

‐80

‐70

‐60

‐50

‐40

‐30

‐20

‐10

0

10

20

‐80 ‐70 ‐60 ‐50 ‐40 ‐30 ‐20 ‐10 0 10

% Cha

nge CO

LLEC

TIVE

BAR

GAINING

COVE

R, 200

0‐20

16

% Change UNION DENSITY, 2000‐2016

% c

hang

e in

col

lect

ive

barg

aini

ng c

over

age,

20

00–2

016

% change in union density, 2000–2016


29

Figure 5 shows how the distribution of collective bargaining coverage evolved from 2000 to 2016. The figure confirms a continuous decline over the 17 years, with the excep-tion of the crisis years, which show some resilience in some countries, before a collapse in 2010–2011 under the influence of fiscal consolidation programmes in several coun-tries. Between 2000 and 2016 the median value decreased, while the dispersion widened, with an elongated rectangle in 2016 compared with 2000: greater dispersion with lower median would be consistent with the scenario of ‘downward convergence’ (or Scenario 4) on industrial relations among the EU28. But how does this compare with the evolution of the LSSI as described in Figure 1?

One of the key elements of Figure 1 was the dynamics of the weighted median for each of the years for which we estimated a box-plot. The median marks a year-to-year distinction between countries among the EU28 whose LSSI is above or below; that is, countries – and the population in Europe – that on average are below or above the sum-mary measure that indicates the goals of the EPSR.

Figure 6 is our final figure and shows the weighted average in terms of collective bar-gaining coverage among countries (and the respective population) with an LSSI above the median versus countries (and their populations) with LSSI below the median. The fig-ure is quite telling: countries with LSSI below the median (their labour market and social dynamics indicator are below the EU28 weighted median) consistently show a lower pro-portion of workers covered by collective bargaining compared with countries with labour market and social dynamics indicators, which are above the EU28 median. In the period following the global financial crisis countries, which registered more solid labour market and social conditions (LSSI above the median), also showed a return to higher values of collective bargaining coverage despite a 10 per cent drop during the period 2006–2007: from then onwards, collective bargaining coverage was maintained almost constantly at

Figure 4 Collective bargaining coverage (2016) and changes in trade unionisation density, by country and EU28, 2000–2016

Note: ILO estimates.

AT BE

BG

HR

CYCZ

DE

DK

EE

FI

FR

GR

HU IE

IT

LVLT

LUMT

NL

PL

PT

RO

SK

SIES

SE

GB

EU28

0

10

20

30

40

50

60

70

80

90

100

110

-80 -70 -60 -50 -40 -30 -20 -10 0 10

Col

lect

ive

barg

aini

ng c

over

age,

201

6

% Change union density, 2000-2016


30

about 80 per cent. Among countries with an LSSI below the EU28 weighted median, col-lective coverage declined gradually as of 2009. At the end of the period, collective cover-age among the LSSI group below the EU28 median was 17.5 per cent lower compared with the year 2000. In sum, Figure 6 would suggest that stronger industrial relations, as summarised and measured by collective bargaining coverage, is not conducive to lower labour market dynamics and socio-economic related outcomes: instead, the evidence can-not reject that stronger industrial relations can be related to the positive outcomes that the European Union aims at promoting through the principles and rights defined by the European Pillar of Social Rights.

The national chapters in this volume further document this relationship. They seem to confirm that the countries that have experienced some weakening of their industrial relations systems during the crisis (Greece, Portugal, Spain) have also experienced the most substantial decline in their labour market and social indicators. While countries such as Greece, Spain and Portugal were on the way to convergence towards the EU28 average, they seem to have started diverging from the crisis onwards. This process of divergence seems to be prolonged, as explained in the chapter on Portugal. Slovenia, a country known for its strong industrial relations system – in particular at tripartite level – has experienced a significant weakening of industrial relations mechanisms over recent years (see chapter on Slovenia), and has undergone a deterioration of its LSSI (from tenth to thirteenth place).

The crisis seems thus to have brought about some divergence in industrial relations in the EU. Mechanisms of bargaining coordination, notably on wages, seem to have been interrupted in the countries hit hardest by the recession. As a result there seems to be a sort of divide between a group of countries with more cohesive and coordinated indus-trial relations and wage bargaining institutions, as well as lower inequality levels that are reflected in a better LSSI, and another group of countries in which industrial relations are more uncoordinated, and income equality and LSSI levels are lower. At the same time, a number of central and eastern European countries have greatly improved their

Figure 5 Further training in companies with and without works councils, Germany

Source: Cantner et al., 2014.

Figure 5 Changes in the collective bargaining coverage rate, EU28, 2000–2016 (all individual years)

Note: ILO own estimates.

0

10

20

30

40

50

60

70

80

90

100

Col

lect

ive

Barg

anin

g co

vera

ge

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016


31

LSSI alongside economic growth, but without developing their industrial relations sys-tems much. While those countries have benefitted from ‘newcomers’ growth, with sud-den access to the whole EU market and also to some EC funding, they may have more difficulties maintaining their current growth rate in the future. The dramatic reduction of industrial relations in these countries may render their progress with regard to the LSSI more difficult.

5. CONCLUSIONS

The analysis proposed in this chapter and the national chapters seems to lead to three important conclusions.

First, an analysis of the convergence of EU countries over the past 16 years shows that, after an upward convergence scenario until 2007, the crisis has changed the picture and led to rapidly increased dispersion, so that over this long-term period (2000–2016), EU countries appear to have diverged rather than converged on the fulfilment of socio-economic indicators. We have tried to summarise trends in social and labour market in-dicators that are fully relevant for the achievement of the European Social Pillar rights. The results show a greater divergence between countries. This is a worrying trend, even if some more encouraging progress is reported in recent years in the national chapters in this volume.

The second conclusion is that such erosion of social indicators has also been extended to industrial relations with a general decline in collective bargaining coverage and trade unionisation. In many countries, the fall in those two indicators is above 50 per cent over 17 years. A certain divide with regard to industrial relations can also be observed , be-tween a group of countries with more resilient industrial relations and lower inequality levels, and another group of countries in which industrial relations have been weakened,

Figure 6 Evolution of collective bargaining coverage, EU28 by LSSI distribution, 1999–2017

20

30

40

50

60

70

80

90

1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

% C

olle

ctiv

e B

arga

inin

g C

over

age

CBC average in countries with LSSI below the EU28 media

CBC average in countries with LSSI at or above the EU media


32

and income equality increased. A number of newer EU members are also continuing their economic growth without developing industrial relations systems very much. This process may undermine the sustainability of their growth in the future, as shown by the severe adjustments they had to absorb during the crisis.

Third, this volume shows that the resilience of industrial relations has often coincided with improved performance and better resilience to the crisis. Our analysis in this chapter confirms in fact that stronger industrial relations, as summarised and measured by collec-tive bargaining coverage, is not conducive to lower (a decline in) labour market dynam-ics and socio-economic related outcomes, but can, on the contrary, be related to positive outcomes with regard to the European Pillar of Social Rights. The chapters in this volume seem to confirm this through more detailed evidence. Collective bargaining in particular has helped to negotiate alternatives to layoffs and assisted many countries in maintaining their level of labour and social indicators. Some collective agreements presented in this volume succeeded in reducing inequalities with regard to access to jobs, social protection and pensions, and to better protect workers confined in part-time employment, temporary contracts, self-employment or in new forms of jobs and working arrangements. Some other agreements have permitted employees to achieve a better balance between work and family life. Higher levels of trade unionisation and collective bargaining coverage of women seem to have had a positive effect on narrowing the gender pay gap.

The development of national industrial relations institutions, such as minimum-wage setting or tripartite pacts was also found to play a role, especially when the strength of the social partners and autonomous collective bargaining had been eroded. The participation of the social partners in vocational training – as in Germany, Sweden and Belgium – or in company-based training (as in France) has also helped in facilitating access to good quality jobs.

Another conclusion from the analysis presented in this chapter is that better measure-ment of convergence with regard to European Social Pillar rights might be needed. We saw the value of the tools and analysis proposed in this chapter as a basis for constructing a solid and comprehensive Summary Indicator that would help in capturing the trends on indicators related to the EPSR. We hope this volume will stimulate further research in that direction.

To sum up, the different chapters try to respond to the main question of this volume: are industrial relations helping EU countries in their pursuit of social and economic con-vergence? If the impact of different elements of industrial relations on socio-economic convergence indicators is somehow confirmed through the reading of this volume – while the European Commission in its social pillar agenda concludes that such social indica-tors definitely contribute to economic growth and convergence – then industrial relations should definitely be treated as a key lever for achieving cross-country convergence.

In this respect we hope that this volume may help to complement policymakers’ views and the options available to them. And that rather than weakening or dismantling indus-trial relations as obstacles or constraints on the road to convergence, they may opt to maintain and even strengthen such institutions so that they can release their full potential for sustainable growth and a return to convergence.


33

REFERENCES

Alcidi, C., Ferrer, J.N., Di Salvo, M., Musmeci, R. and Pilati, M. (2018) Income Convergence in the EU: A tale of two speeds, Brussels: CEPS.

Atkinson, A.B. (2015) Inequality. What Can Be Done? Oxford: Oxford University Press. Crouch, C. (2013) Making Capitalism Fit for Society. Polity Press, Cambridge.De Spiegelaere, S. and Jagodzinski, R. (2015) European Works Councils and SE Works councils

in 2015. Facts and figures, Brussels: ETUI.EPSC (2015) The social dimension of economic and monetary union – Towards convergence

and resilience, EPSC Strategic Notes, Issue 5, 18 May 2015, Brussels, European Political Strategy Centre EC.

Eurofound (2015) Pay in Europe in different wage-bargaining regimes. Luxembourg: Publica-tions Office of the European Union.

Eurofound (2017) Converging economies, diverging societies? Upward convergence in the EU, Eurofound, Dublin. Available at: https://www.eurofound.europa.eu/fr/events/foundation-forum-2017; https://www.eurofound.europa.eu/fr/publications/other/2017/converging-econo-mies-diverging-societies-upward-convergence-in-the-eu-foundation-forum-2017

European Commission (2017a) Commission Recommendation on the European Pillar of Social Rights, C(2017)2600 final of 26 April 2017, Brussels.

European Commission (2017b) Establishing a European Pillar of Social Rights, COM(2017)250 final of 26 April 2017, Brussels

European Commission (2017c) Reflection paper on the deepening of the Economic and Mon-etary Union, COM(2017)291 of 26 April 2017, Brussels.

Grimshaw, D., Koukiadaki, A. and Tavora, I. (2017) Social Dialogue and Economic Perfor-mance – What matters for business – A review, Conditions of Work and Employment Series No. 89, INWORK, ILO, Geneva.

ILO (2016a) Building a social pillar for European convergence, Studies on growth with equity, Geneva: ILO.

ILO (2016b) Evaluating the effects of the structural labour market reforms on collective bar-gaining in Greece, Conditions of work and employment series No. 85, INWORK, Geneva.

Koll, W. (2013) The new economic governance arrangements and autonomous collective bar-gaining in the European Union. Dialogue-based macroeconomic coordination. IMK Study.

Papadakis, K. (ed.) (2011) Shaping Global Industrial Relations: The impact of International Framework Agreements, Geneva: ILO.

Sapir, A. and Wolff, G.B. (2015) What to reform and how to do it?, Bruegel Policy Brief, Febru-ary 2015.

Vandenbroucke, F. (2013) ‘A European Social Union: Why we need it, what it means’, Rivista Italiana di Politiche Pubbliche, 2 (13): 221–47.

Vandenbroucke, F. (2017) Structural convergence versus systems competition: Limits to the diversity of labour market policies in the EMU, Discussion Paper 065, July 2017, European Commission, Brussels.

Vaughan-Whitehead, D (ed.) (2018) Reducing inequalities in Europe – How industrial relations and labour policies can close the gap, Edward Elgar in cooperation with the ILO, Geneva.

Visser, J. (2016). What happened to collective bargaining during the great recession? IZA Jour-nal of Labor Policy, 5: 9.

Watt, A. (2012) Austerity is not the solution. Contributions to European Economic Policy, FEPS, Brussels, pp. 97–118.


34

ANNEX 1

LSSI: Labour and Social Summary IndicatorAreas Indicators Notes and Treatment SourceAccess to labour market • Employment rate

• Unemployment rate, overall

• Youth unemployment rate

• Long-term unem-ployed

Employment and unem-ployment rates based on the population age 15 to 64. Youth and long-term unemployed based on total unemployed

EUROSTAT – based on LFS and EU-SILC estimates

OECD Stats

Access to education and skills formation

• Proportion of depend-ent workers in educa-tion/skills enhancing activities

• Proportion of work-force in training with formal education below secondary

• Proportion of work-force in training with formal education equivalent to second-ary and vocational

• Proportion of work-force in training with formal higher educa-tion (university and above)

All four indicators based on the fraction of dependent workers that claimed to be in train-ing/skills formation ‘at some point during the last 4 weeks at the time of the survey’

EUROSTAT – based on LFS

Access to decent earnings

• Real wages in PPP 2011 US$

• Real wage growth• Kaitz index

The inclusion of real wages in PPP terms helps to identify ‘decent wages’ whereas the growth rate helps to identify the relation between wages and the economy. The Kaitz for countries without a minimum wage is ap-proximated by imputing the value of net income received by minimum income benefit receivers (single person, includ-ing supplements) as % of median income.

OECD StatsWDI – World develop-ment Indicators (World Bank)ILO Global Wage Da-tabase

Access to stable em-ployment (standard forms of employment)

• Average number of hours worked per week by wage em-ployees

• Percentage of part-time among wage employees

• Percentage of in-voluntary part time among part time workers

Part-time and contrac-tual conditions based on national definitions (as opposed to the construc-tion of the indicators using a standardised common definition)

EUROSTAT – based on LFS


35

LSSI: Labour and Social Summary Indicator (cont.)Areas Indicators Notes and Treatment SourceExpenditure on social protection

• Direct social protec-tion expenditure per capita

• Other social protec-tion expenditure per capita

Social protection benefit expenditure are trans-fers to households, in cash or in kind, that include disability, sick-ness/healthcare, old age, survivors, family/chil-dren, unemployment, housing and social exclusion. Other social protection expenditure includes administra-tion costs and other expenditure (that may be exclusive of coun-tries) associated with relieve transfers and not included in the first indicator. The two in-dicators together cover ‘total expenditure on social protection’.

EUROSTAT – National accounts

Inequality • Share of total dis-posable household income of households at the bottom decile

• Share of total dis-posable household income of households at the top decile

• Total share of house-holds income from households below the median as ratio to the share accrued by households above the median (excluding top and bottom deciles)

All indicators area based on total house-hold disposable income. The 3rd variable is a ratio of two shares: the numerator is the addi-tion of the share accrued by households in the 2nd to 5th decile of the income distribution; the denominator is the share accrued by households in the 6th to the 9th deciles.

World Income Inequal-ity Database (WIID-UNU-WIDER), Release January 2017

Economic performance and distributional out-comes

• Unadjusted labour income share

• Growth in real GDP per capita

Unadjusted labour income shares excludes own account workers which is a relative small fraction of the work-force for countries in EU28 group

LIS : AMECO (EU-Commission)GDP: World Develop-ment Indicators (WDI-WB)


36

Industrial Relations Areas Indicators Notes and Treatment SourceUnion density • Union density rate Union density rate, net

union membership as a proportion of wage earners in employment

ICTWSS – AIAS, Release 5.1 (September 2016)

Collective bargaining coverage

• Proportion of employ-ees with the right to collective bargaining

Collective bargaining coverage rate corre-sponds to the ratio of employees covered by collective agreement, divided by all wage earners (dependent workers) with right to bargaining.

ICTWSS – AIAS, Release 5.1 (September 2016)


37

ANNEX 2 SUMMARISING LABOUR MARKET AND SOCIO-ECONOMIC INDICATORS INTO A CONVERGENCE MEASURE, THE LABOUR AND SOCIAL SUMMARY INDICATOR (LSSI)

Comparing countries using several indicators can be a complex task. For example, a comparative labour market analysis of the EU28 using employment rates, unemploy-ment rates and involuntary part-time would involve reviewing and assessing at least 84 tendencies. The complexity increases when more than three indicators are involved, especially if these are not independent. A solution to this complexity is to construct a measure in which the indicators are summarised into one single variable: this technique is what lies behind the construction of well-known indexes such as the Human Develop-ment Index (HDI, UNDP) or the Better Life Index (BLI, OECD). In practice, the tech-nique involves constructing a number (a score) for each of the countries involved (or units, or regions) so that these can be compared with each other in equal terms.

Constructing a summary measure relies on two basic steps. The first consists of con-verting each indicator into the same unit of measurement, that is, normalising each in-dictor into a range of values (say, from 0 to 1), where going from one extreme value to the other (say, from scoring 0 to scoring 1) implies going from lower (less desired) to higher or more desired outcomes. For example, if we have 10 countries with employ-ment rates ranging from 35% to 70%, the country with 35% could be attributed a value of 0 and the country with 70% a value of 1. The other eight countries would be attached values higher than zero but less than one, where such values (scores) would reflect the within-group relative distance in employment rates. The trick with the normalisation of an indicator is an adequate strategy that assigns each intra-range unit a value (between 0 and 1) that captures the relative distance between countries. The strategy for construct-ing the Labour and Social Summary Indicator (LSSI) considers the need to normalise indicators so that the scores retain the time-series characteristics of the original indica-tors: this allows us to compare the relative performance of countries at a particular point in time, and the between-country and within-country performance over time, which is required for measuring convergence. To describe the method let us assume we have data for a sequence of years (for example, 2000–2016, so that the first period is the year 2000) and 10 countries and let’s take the employment rate as illustration. In the year 2000, the highest and lowest employment rate among the 10 countries is 35% and 70%, so these two values become the lowest and highest base year normalising values, namely,

and . Applying expression (1) to the employment rate of each country at time t, that is, , implies attaching a score to each country in the set:

200035m

200070m ,c tv

,c tnv

, 0,

0 0

c t tc t

t t

v mnv

m m

For example, if country Z among the set of 10 has an employment rate of 49% in the year 2000, and its normalised score in 2000 will be For subsequent years (that is, from 2001 to 2016) we keep the same normalising values – in general, – to normalise the employment rate of the same group of countries. In the con-struction of indexes such as the HDI or the BLI, the normalising parameters change at each point in time – that is, they become – so that the scores are only valid to ‘rank’ countries at a particular point in time. With the normalising strategy in (1), the score values attached to each country retain their cross-section nature (that is, this can still be used as a rank or a comparison measure between countries) but anchoring the normalisation to some base year means that the scores attached to the indicators that are

, 200049%

Zv

, 20000.4

Znv

0 0( ),

t tm m

( ),t t

m m

38


normalised over subsequent years gain a time-series dimension: comparing the time movements of the score for a single country shows growth, while comparing the scores between a group of countries shows relative performance between countries that grow or shrink with respect to that particular indicator(s). It can be shown that the properties of the normalised variable using the standard method – – are retained when compared with the modified method in (1). The only condition is that the value of the original indicators cannot grow unbounded, while the denominator in (1) has to be positive. It is rare to find a situation where these two conditions are not met. Finally, one of the conditions for normalising indicators is that once these are normalised all of them follow the same directional interpretation (low values imply less desired out-comes, higher values more desired outcomes). In the case of employment rates ‘higher is better’ would be a reasonable desire, but in the case of unemployment, for example, the opposite is true. When the nature of indicators is such that ‘lower is a more desired outcome’, then the normalisation of the indicator consists of subtracting 1 minus the result obtained when applying (1). That is, apply [ ].

The second and final step consists of joining all the normalised indicators into one single measure. This can be done by taking a simple average between the score values of all the normalised indicators, for each country and at each point in time. For example, if we have 28 countries and 20 indicators, add the score values of the 20 normalised indicators for each country, and divide these by 20. The result is an indicator of dimen-sion 1x28.

With the above reading of the index, it becomes clear that the treatment of the so-called ‘controversial indicators’ is such that countries with the highest value (for ex-ample, with the highest permanent employment rate) will receive a higher score from this particular indicator. However, the construction of the LSSI does not imply that countries have to pursuit a maximal or minimal strategy on any particular indicator for EU28 convergence on the European Pillars. Instead, the analysis is such that at, any given point in time, countries are scattered over a range of values as a result of their underlying indicators. Identifying convergence involves examining the changes in the distribution of the LSSI values over a sufficient time span (in the case of this chapter this amounts to the period 2000 to 2016) comparing such movements to the first year of the analysis (in this case, the year 2000). The slight modification in the construction of the LSSI relative to the construction of measures such as the Human Development Index or the Better Life Index – to mention only a couple – provides the LSSI with a time-series dimension that validates the use of the measure as one of convergence or divergence from some initial point in time.

The distributional changes of the LSSI can also shed light on the relation between industrial relations, and its evolution, and the evolution of labour market and social dynamics. This in fact completes the circle in Chapter 1 because the evolution of in-dustrial relations is at the heart of each chapter in this volume. Again, trying to com-pare 20 independent indicators with those frequently used to capture industrial relations (for example, collective bargaining coverage and union density) could in fact bring too much complexity into the analysis, while using the LSSI reduces the problem to a single dimension.

( ),t t

m m

,1 c tnv

1. Convergence in the EU: what role for industrial relations? · the impact of industrial relations...

Documents

Transcript of 1. Convergence in the EU: what role for industrial relations? · the impact of industrial relations...